On 21/10/22 16:16, Andy Shevchenko wrote:
> On Fri, Oct 21, 2022 at 01:19:26PM +0100, Valentin Schneider wrote:
>> The recently introduced sched_numa_hop_mask() exposes cpumasks of CPUs
>> reachable within a given distance budget, wrap the logic for iterating over
>> all (distance, mask) values inside an iterator macro.
>
> ...
>
>> #ifdef CONFIG_NUMA
>> -extern const struct cpumask *sched_numa_hop_mask(int node, int hops);
>> +extern const struct cpumask *sched_numa_hop_mask(unsigned int node, unsigned int hops);
>> #else
>> -static inline const struct cpumask *sched_numa_hop_mask(int node, int hops)
>> +static inline const struct cpumask *
>> +sched_numa_hop_mask(unsigned int node, unsigned int hops)
>> {
>> - if (node == NUMA_NO_NODE && !hops)
>> - return cpu_online_mask;
>> -
>> return ERR_PTR(-EOPNOTSUPP);
>> }
>> #endif /* CONFIG_NUMA */
>
> I didn't get how the above two changes are related to the 3rd one which
> introduces a for_each type of macro.
>
> If you need change int --> unsigned int, perhaps it can be done in a separate
> patch.
>
> The change inside inliner I dunno about. Not an expert.
>
That's a rebase fail, this should all be in the first patch, my bad.
